Service as Software
One of the biggest benefactors of the AI boom was Accenture who in 2024 booked over $3 billion in revenue from generative AI alone. Just as we saw with enterprise cloud transformation, medium and large enterprises need hand holding in adopting new technology. We’re inspired by our portfolio company QA Wolf who scaled quickly from 0 to 8-figures in ARR using a service-led GTM strategy.
Starting with co-pilot tools and moving towards autonomous agents, there’s a powerful ROI narrative which will lead to leaner, smaller teams in every department and white-collar professional type. We’re looking for product-oriented founders who are customer obsessed and won’t shy away from the complexities of a service-led growth strategy in delivering software primitives.
Sensors & Science
As AI accelerates the commoditization of software, the rate limiter in product innovation is not code, but new ways to collect and use proprietary data. With hardware and synthetic components becoming more affordable, we’re interested in telemetry startups that use sensors and science to outfit the physical world with new software applications.
We’re looking for novel approaches to capturing real-world data and opportunities to build new, programmable chassis. Index Biosystems is an example of a multi-disciplinary startup in our portfolio building a digital twin of the food industry.
Small Businesses
Japan boasts a remarkably diverse ecosystem of small and micro businesses, many of which are founder-owned and operated. While these businesses deliver exceptional consumer experiences, they also face significant challenges, including limited infrastructure, financial instability, and inadequate succession planning, which can result in extreme hardship.
Drawing inspiration from the success of portfolio companies like Grow Therapy and Fora Travel, we are particularly interested in how technology can empower entrepreneurs to break away from top-heavy, investor- or PE-owned organizations.
We believe that software combined with light operating services can build structurally advantaged back offices for scaling high-quality micro businesses. This model not only improves the economic balance between owners and operators but also preserves the operators’ autonomy and enables them to focus on their craftsmanship.
Vertical AI
Here are three factors we consider when assessing prospective investments:
- Outperformance: We look for startups that can articulate significant advantages of last-mile fine-tuning over generalized models and non-AI incumbents. Opinionated industry templates, ecosystem integrations, and customer data moats can all provide sufficient differentiation while building longer term defensibility.
- Labor Displacement: High labor costs and employee turnover significantly bolster the perceived ROI of AI investments. We prefer startups that can fully displace FTEs, even if they begin as co-pilot SaaS or human-in-the-loop services. We’re particularly interested in use cases where AI enables continuous, broader, and/or better coverage than human operators.
- Adoption Tailwinds: Regulatory pressure, surging customer demand, and talent shortages are accelerants for AI adoption. In addition to exogenous tailwinds, we’re particularly excited by startups that are capable of generating new revenue for customers leveraging AI.
Healthcare Fintech
Even before the tragic assassination of UHC’s CEO, most people were aware of the tension between payors and patients. Left to their own devices, the dissonance has widened between those who pay for healthcare (insurance carriers, self-funded employers, government), those who provide services (providers and staff), and the patients that receive care.
Garner Health, TPIC, Otter Health are examples of portfolio companies who is re-aligning financial incentives and hold all stakeholders accountable. We believe payments is the ultimate meritocratic ledger in which we reduce waste and improve patient outcomes in healthcare.
Computational Sciences
The transformative potential of computational chemistry and synthetic biology (ex. in silico therapeutic development) is uniting historically siloed STEM disciplines. In our early explorations of this field, we passed on several startups due to an anticipated adoption gap caused by fragmented expertise among lab scientists, data scientists, DevOps engineers, and software developers. There are far too many siloed specialists in an industry that’s increasingly reliant on multi-disciplinary collaboration. Change is inevitable.
We are particularly interested in software and services that promote collaboration across disciplines, simplify complex skill gaps, and creating unique marketplace / network effects. These solutions are especially valuable at the intersection of pharmaceutical companies, manufacturers, suppliers, and downstream customers.